Introducing DeepTutor v1.5: Agent-native Personalized Tutoring. Our core belief: Tutoring should be a data loop, not disconnected features. DeepTutor connects teaching, practice, behavioral traces, unified Runtime, inspectable memory, and proactive IM companions into one evolving learner model. DeepTutor is an agent-native learning workspace that connects tutoring, problem solving, quiz generation, research, visualization, and mastery practice in one extensible system. ✨ Key Features of DeepTutor v1.5 1/ - One runtime for every mode — Chat, Quiz, Research, Visualize, Solve, and Mastery Path all run on the same agent loop. You switch the objective, not the engine, and context moves with you. 2/ - Connected learning context — Knowledge bases, books, Co-Writer drafts, notebooks, question banks, personas, and Memory stay available across every workflow instead of living in isolated tools. 3/ - Subagents and Partners — consult a live Claude Code, Codex, or Partner from any turn, import their past conversations, and run persistent IM companions on the same brain. 4/ - Multi-engine knowledge — versioned RAG libraries across LlamaIndex, PageIndex, GraphRAG, LightRAG, or a linked Obsidian vault, with pluggable document parsing. 5/ - Extensible tools and skills — built-in tools, MCP servers, image / video / voice generation models, and installable community skills from EduHub. 6/ - Inspectable memory — L1 traces, L2 surface summaries, and L3 synthesis make personalization visible and editable, with a Memory Graph that traces every claim back to its evidence. GitHub: Website: Paper: